With the discovery that mutations in fourteen different keratin genes can cause eight distinct disorders that exhibit epithelial fragility, comes the realization that the keratin filament network is crucial to the structural integrity of epithelial tissues exposed to mechanical stress. The identification of mutations causing these disorders has improved our understanding of keratin intermediate filament structure and function, particularly with regard to the highly conserved regions of the central rod domain. However, two central questions in keratin biology still remain: First, how are the individual keratin genes regulated during normal and abnormal development? To address this question, we have chosen to look at two related keratins that have widely different regulatory properties. These are keratins K1 and K6. K1 is one of the first markers of keratinocyte differentiation, with its expression initially detected in post-mitotic basal cells. Moreover, defects in K1 have been identified as causative in the skin disease epidermolytic hyperkeratosis, EHK. There are at least five mouse K6 genes, and specific isoforms are differentially expressed in the outer root sheath of hair follicles and in other epithelia. Interfollicular expression of K6 only occurs when keratinocytes are stressed or perturbed, such as in wounding. Mutations in two human K6 genes have been detected in different forms of pachyonychia congenita. The functional significance of multiple K6 genes remains to be determined. To address this question, we have initiated knock out experiments and on the basis of data generated to date, we suggest that certain human K6 genes have not yet been identified, and are in fact candidates for mutations in other disorders. The second question concerns the roles of the keratin and C-terminal end domains. These sequences are distinctive for each keratin protein but are remarkably well conserved across species and presumably have functional significance. In vitro studies have suggested that these domains interact with desmosomes and/or the cell envelope, however these interactions have not been confirmed by in vivo models. Two transgenic approaches are proposed to determine the functional role of the end domains in vivo. Finally, dominant skin disorders which affect post-mitotic epidermal cells present a difficult challenge for gene therapy. We propose to develop a mouse model for one of these disorders, EHK, and use it to assess gene therapy approaches.